Reexamining Nonlinear Structural Equation Modeling Procedures: The Effect of Parallel and Congeneric Measures

The current study examines the performance of the extended unconstrained approach (EXUC) and the latent moderated structural equation modeling procedure (LMS) in situations where quadratic and interaction terms are tested simultaneously and investigates their limitations with regard to the employment of parallel and congeneric measures, relatively low indicator reliabilities, and relatively large numbers of indicators. By means of a Monte Carlo study, we found LMS to be the best option for testing multiple nonlinear effects given sufficient sample size (n ≥ 500) and normally distributed exogenous variables. Its advantages became more prominent when indicator reliabilities were heterogeneous and small. The EXUC was a viable option for estimating the model when indicators were parallel and exhibited large indicator reliabilities. An empirical example of the results is provided, and the relevance of measurement model characteristics to assess nonlinear relationships is discussed.     

The relationship between attributions and depression varies across attributional dimensions and across samples

Two hypotheses implicit in the use of composite measures of attributions in tests of learned helplessness theory (but not implicit in the theory itself) were tested: the hypotheses that relationships between depression and the three types of attributions are equal in magnitude and are linear. To test these hypotheses, data from three published studies of the reformulated learned helplessness theory (Abramson, Seligman, & Teasdale, 1978) were reanalyzed. The hypothesis that internal, stable, and global attributions are equally related to depression was tested and rejected. Increases in internal attributions were related to depression in one sample; increases in global attributions for negative events were related to depression in two samples; stability attributions for negative events were unrelated to depression. The relationship between attributions and depression was nonlinear in one of the three populations studied. Finally, a third hypothesis was tested: the hypothesis that the relationship between attributions and depression is equal across samples. The hypothesis was rejected: attributions for negative events were more highly related to depression in a psychiatric sample than in normal populations. Implications of these findings for learned helplessness theory and for the use of composite measures of attributional style are discussed.     

Religiosity and Premarital Sexual Behaviors Among Adolescents: An Analysis of Functional Form

While numerous studies show a persistent inverse association between religion variables and adolescents’ sexual behaviors, the nature of this relationship is not well understood. Specifically, many previous studies presuppose that the associations between adolescent religiosity and sexual behaviors are linear. However, a number of studies have also identified important nonlinearities of religious influence during adolescence, with highly religious individuals being distinct from their peers. Incorporating this knowledge into a theoretically motivated modeling approach, this article provides a comparative analysis of functional forms describing the relationships between religiosity and adolescent sexual behaviors. Using data from two waves of the National Study of Youth and Religion, a linear functional form is compared with nonlinear alternatives that link multiple religion measures to outcomes of sexual activity. Results show that the majority of these relationships are best defined by nonlinear functional forms, suggesting that the influence of religiosity increases as individuals become more religious.     

Validity of the strategic management simulations to predict real-world productivity

The performance of individual businesses and the economy as a whole hinges on the productivity of human capital, yet measuring productivity at an individual level has proved difficult. The Strategic Management Simulation (SMS) methodology comprises several decision-making assessment tools that have been used for several decades to estimate real-world productivity. Performance on productivity measures using a shorter version of the SMS methodology was gathered from a sample of 111 knowledge workers from major corporations in the U.S. and compared with their income level. In linear models, every percentile increase in SMS performance was associated with a $791–$1,113 increase in income in each of the four domains of decision-making performance provided by the SMS tool. The models had R²’s ranging from 0.65 to 0.75, indicating high levels of correlation between SMS performance and income levels. This study is the first to show linear relationships between SMS performance and other indicators of productivity, indicating that the SMS is a valid predictor of productivity and that SMS scores can be used to predict job satisfaction and performance.     

The use of artificial neural networks and multiple linear regression in modelling work–health relationships: Translating theory into analytical practice

Although psychological theory acknowledges the existence of complex systems and the importance of nonlinear effects, linear statistical models have been traditionally used to examine relationships between environmental stimuli and outcomes. The way we analyse these relationships does not seem to reflect the way we conceptualize them. The present study investigated the application of connectionism (artificial neural networks) to modelling the relationships between work characteristics and employee health by comparing it with a more conventional statistical linear approach (multiple linear regression) on a sample of 1003 individuals in employment. Comparisons of performance metrics indicated differences in model fit, with neural networks to some extent outperforming the linear regression models, such that R ² for worn-out and job satisfaction were significantly higher in the neural networks. Most importantly, comparisons revealed that the predictors in the two approaches differed in their relative importance for predicting outcomes. The improvement is attributed to the ability of the neural networks to model complex nonlinear relationships. Being unconstrained by assumptions of linearity, they can provide a better approximation of such psychosocial phenomena. Nonlinear approaches are often better fitted for purpose, as they conform to the need for correspondence between theory, method, and data.     

The Use of Priority Information in Performance Feedback

To provide the most value to an organization, individuals must focus efforts on those outputs that have maximum value to the organization. However, the realities of the workplace are that information about performance and how individuals can add value to an organization is complex; that is, determining which activities to focus on is a fairly complex task because such information is nonlinear, whereby the importance of a certain activity may change depending on the level of performance at any given time. This article empirically examines whether providing complex, nonlinear feedback that reflects the complexities of the real world will result in a change in work outputs that reflect these nonlinearities. This article assesses whether employees can and do use complex feedback that reflects the complexity of the real world in the intended ways. In addition, this article examines whether individuals use feedback that includes priority information beyond feedback that reports on simple relative importance. Data were analyzed from 9 work units for time periods averaging a little over a year. Results indicate that, as predicted, when individuals have access to both complex, nonlinear priority information and simple, linear relative importance information, individuals can and do use the complex data in the intended way.     

Effects of Nonlinearity and Heteroscedasticity on the Validity of Conscientiousness in Predicting Overall Job Performance

Most research on the relationship between personality and overall job performance assumes linearity and homoscedasticity. This study investigated the prevalence and nature of nonlinearities and heteroscedasticies in relationships between conscientiousness and supervisory ratings of overall job performance across five independent samples using both concurrent (k = 4) and predictive (k = 1) designs. Hierarchical polynomial regression analyses found evidence of robust linear effects but no evidence of statistically significant quadratic or cubic effects. A statistically and practically significant heteroscedastic effect was found in only one sample such that lower errors of prediction were evidenced in the ends in comparison to the middle of the bivariate distribution. Implications for the use of conscientiousness in personnel selection are suggested. Limitations of the current study and directions for future research are noted.     

Linearity of Personality–Performance Relationships: A large‐scale examination

This paper reviews theory and past research concerning nonlinear relationships between personality tests and job performance. We compare linear and nonlinear regression models for each of the 32 Occupational Personality Questionnaire (OPQ32i) scales using job performance as the criterion variable. We evaluate linear, quadratic, and cubic relationships between personality and job performance by calculating changes in R (ΔR) between the linear and nonlinear models. We use 3 decision rules of varying levels for ΔR and divide a large sample randomly into 2 samples to cross‐validate any nonlinear effects found. Few scales exhibited nonlinearity and the magnitude of the departures from linearity were small.     

To Be,or Not to Be,Linear: An expanded representation of personality and its relationship to leadership performance

In this study, we tested for a set of complex, nonlinear relationships between derailing/dark side personality composites and leadership performance using two independent samples of managers/leaders (N=1306 and N=290 for Study 1 and 2, respectively). Based on the structure and characteristics of the derailing/dark side trait composites, we expected the relationship between these composites and leadership performance would best be described with an inverted U function. In Study 1, we found evidence of a nonlinear relationship between the derailing composite, as measured by the Global Personality Inventory^©, and behaviorally based, structured ratings of leadership performance in an assessment center. Similarly, Study 2 found evidence for a nonlinear relationship between dark side composites, as measured by the Hogan Development Survey^©, and supervisory/other ratings of leadership performance. We discuss the implications of these complex, nonlinear findings with respect to the continued use of personality for the selection and promotion of future leaders. Limitations of the current study and directions for future research are noted.     

The relationship of novelty preferences during infancy to later intelligence and later recognition memory

In the present study, a group of children originally tested for visual novelty preferences at 7 months and seen for estimates of intelligence at 3 years were revisited and tested for intellectual functioning and for visual recognition performance at 5 years. Prediction from performance at 7 months to intellectual functioning at 5 years was significant and remained at the same level as prediction from 7 months to 3 years (r = .42 at each point). In addition, it was shown that data on early novelty preferences could be combined with information on birth-order and parental education to achieve good prediction (R = .65) of five-year IQ scores. A further goal of the present study was to discover if infants' preferences for visual novelty, assumed to be an early measure of visual recognition memory would be more highly associated with later visual recognition performance or with later intellectual level. The results indicate that novelty preferences were more highly related to later intelligence quotients than to later recognition performance and such a pattern of relationships did not to be artificial.     

Post-exercise pulse pressure is a better predictor of executive function than pre-exercise pulse pressure in cognitively normal older adults

Exercise “stress tests” are widely used to assess cardiovascular function and to detect abnormalities. In line with the view of exercise as a stressor, the present study examined the relationship between cognitive function and cardiovascular activity before and after light physical exercise in a sample of 84 non-demented community-dwelling older adults. Based on known relationships between hypertension, executive function and cerebral white matter changes, we hypothesized that greater post-exercise reactivity, as indexed by higher pulse pressure, would be more related to worse performance on frontal-executive tasks than pre-exercise physiologic measures. All participants were administered a comprehensive neuropsychological battery and underwent a Six Minute Walk Test (6MWT), with blood pressure (BP) measures obtained immediately before and after the walk. Pulse pressure (PP) was derived from BP as an indicator of vascular auto-regulation and composite scores were computed for each cognitive domain assessed. As predicted, worse executive function scores exhibited a stronger relationship with post-exercise PP than pre-exercise PP. Results suggest that PP following system stress in the form of walking may be more reflective of the state of vascular integrity and associated executive dysfunction in older adults than baseline physiologic measures.     

Internal Consistency and Power When Comparing Total Scores from Two Groups

Researchers now know that when theoretical reliability increases, power can increase, decrease, or stay the same. However, no analytic research has examined the relationship of power to the most commonly used type of reliability—internal consistency—and the most commonly used measures of internal consistency, coefficient alpha and ICC(A,k). We examine the relationship between the power of independent samples t tests and internal consistency. We explicate the mathematical model upon which researchers usually calculate internal consistency, one in which total scores are calculated as the sum of observed scores on K measures. Using this model, we derive a new formula for effect size to show that power and internal consistency are influenced by many of the same parameters but not always in the same direction. Changing an experiment in one way (e.g., lengthening the measure) is likely to influence multiple parameters simultaneously; thus, there are no simple relationships between such changes and internal consistency or power. If researchers revise measures to increase internal consistency, this might not increase power. To increase power, researchers should increase sample size, select measures that assess areas where group differences are largest, and use more powerful statistical procedures (e.g., ANCOVA).     

Predicting Adolescent Drinking From Alcohol Expectancy Values: A Comparison of Additive,Interactive, and Nonlinear Models1

Additive, interactive, and nonlinear models of alcohol expectancy values were compared using survey data from 1,758 high school students. Expectancies and values independently predicted drinking in the additive model. Expectancies were more important as predictors than were values, and negative expectancies were more important than positive expectancies. Significant expectancy-value interactions also were found. Drinking was highest when positive consequences were believed to be likely and desirable and was lowest when negative consequences were believed to be likely and undesirable. Significant nonlinearities indicated that beliefs about negative consequences had greater effects at lower levels of likelihood and evaluation whereas beliefs about positive consequences had greater effects at higher levels of likelihood and evaluation. However, the interactive and nonlinear effects were small.     

Reliability measures in item response theory: Manifest versus latent correlation functions

For item response theory (IRT) models, which belong to the class of generalized linear or non‐linear mixed models, reliability at the scale of observed scores (i.e., manifest correlation) is more difficult to calculate than latent correlation based reliability, but usually of greater scientific interest. This is not least because it cannot be calculated explicitly when the logit link is used in conjunction with normal random effects. As such, approximations such as Fisher's information coefficient, Cronbach's α, or the latent correlation are calculated, allegedly because it is easy to do so. Cronbach's α has well‐known and serious drawbacks, Fisher's information is not meaningful under certain circumstances, and there is an important but often overlooked difference between latent and manifest correlations. Here, manifest correlation refers to correlation between observed scores, while latent correlation refers to correlation between scores at the latent (e.g., logit or probit) scale. Thus, using one in place of the other can lead to erroneous conclusions. Taylor series based reliability measures, which are based on manifest correlation functions, are derived and a careful comparison of reliability measures based on latent correlations, Fisher's information, and exact reliability is carried out. The latent correlations are virtually always considerably higher than their manifest counterparts, Fisher's information measure shows no coherent behaviour (it is even negative in some cases), while the newly introduced Taylor series based approximations reflect the exact reliability very closely. Comparisons among the various types of correlations, for various IRT models, are made using algebraic expressions, Monte Carlo simulations, and data analysis. Given the light computational burden and the performance of Taylor series based reliability measures, their use is recommended.     

Dispositional and situational moderators of female and male causal attributions

To increase understanding of the relationships between gender and causal attributions, dispositional and situational variables were examined to determine if they affected causal attributions differently if the subjects were females or males. Four dispositional variables—locus of control, neuroticism, achievement motivation, and self-esteem—and five situational variables—expectancy of success, self-reported commitment, perceived productivity, perceived task complexity, and actual performance—were examined for their moderating effects on the gender-causal attributional relationships. The dispositional variables did not moderate any causal relationships. By comparison, four situational variables—performance, commitment, productivity, and complexity—moderated at least one of the gender-attribution relationships. It was concluded that more attention should be directed to the identification of situational moderators.     

Why does a red snake in the grass capture your attention?

It has been suggested that affective stimuli automatically capture attention; this preferential processing is thought to be related to the evolutionary significance of affective stimuli. However, recent evidence suggests that perceptual salience alone might explain why some affective stimuli are more likely to influence attentional processes in certain contexts. In this study, we manipulated affective and perceptual salience to better understand how affective information is processed and how it impacts attentional processes in different contexts. We used stimuli that are both affectively and perceptually salient, while varying the task requirement to encourage the processing of perceptually salient (Experiment 1) or affectively salient (Experiment 2) information. This design made it possible to observe independent and interdependent relationships between perceptual and affective salience. The results showed that when the task encouraged the processing of perceptually salient information, affective salience did not influence task performance. In contrast, when the task encouraged the processing of affectively salient information, affectively salient information impaired task performance. The findings suggest that the affective nature of the stimuli does not always influence attentional processes.     

Prediction of Simulated Infantry-Combat Performance From a General Measure of Individual Aptitude

The relationships between individual soldier aptitudes and the potential for successful combat performance have not been systematically studied. The purpose of this research was to determine the relationships between one measure of individual aptitude, the General Technical (GT) composite from the Armed Services Vocational Aptitude Battery (ASVAB), and performance during simulated infantry combat. Eleven 3-man teams of infantry soldiers defended against enemy rifle squads performing a Movement to Contact, Hasty Attack mission. The correlation between GT and casualties inflicted was r(30) = .37, p < .05; the correlation between GT and casualties sustained was r(30) = -.25, p > .05. Marksmanship scores did not correlate with aptitude or the two criterion measures, partially due to limited variance. Given further validation, these results could provide important information for Army policy and management decisions.     

Evaluating Family Interactions to Inform Exhibit Design: Comparing Three Different Learning Behaviors in a Museum Setting

ABSTRACT

How do we know that families are learning during the museum visit itself? In order to address this question, this study focuses on three frequently evaluated indicators of during-visit learning: time spent, exhibit engagement, and interpretive talk. Each of these types of data makes certain assumptions about family learning and how it occurs. Four hundred ninety-three family interactions at 25 different children's museum exhibits were analyzed to examine what the three measures reveal about family learning behavior. When compared, the three measures were not highly correlated, suggesting that generalizing information about the family experience from a single indicator is insufficient. Looking at all three indicators simultaneously can help evaluators begin to explore the relationships between families’ learning behaviors, and think about how exhibit characteristics can influence potential learning opportunities in informal settings.     

Midsagittal brain shape correlation with intelligence and cognitive performance

Brain shape might influence cognitive performance because of the relationships between functions, spatial organization, and differential volumetric development of cortical areas. Here we analyze the relationships between midsagittal brain shape variation and a set of basic psychological measures. Coordinates in 2D from 102 MRI-scanned young adult human brains were superimposed through a Procrustes approach, and the residual variation was regressed onto 21 cognitive tests performed by the same individuals. Most of the composite and specific variables (including general intelligence, working memory, attention, and executive functions) do not show meaningful correlations with midsagittal brain morphology. However, variables related to mental speed display subtle but consistent correlations with brain shape variation. Such correlations are small, suggesting that the influence of midsagittal brain geometry on individual cognitive performance is negligible. Nevertheless, this evidence can supply information on brain biology and evolution. Areas associated with the parietal cortex appear to be involved in relationships between brain geometry and mental speed. These areas have been associated with relevant endocranial differences between living and extinct humans, and are important as functional and structural components of brain organization. The limited correlation between brain geometry and mental speed among modern individuals might be more relevant when the large paleoneurological variation of the genus Homo is taken into account.     

Procedural Text Structure and Reader Perceptions and Performance

This research tested the relationship between text structural variables (on vs. off the causal chain, active vs. static information, and hierarchical structure) and reader perceptions, reading time, and true-false performance for procedural texts. Twenty-four college-age participants each read 3 procedural texts. As predicted, sentences that were (a) on the causal chain, (b) active, and (c) associated with task steps were read more slowly and judged to be more important (p < .05). The results were similar to those previously found with narratives, except for the hierarchical structure variable. Also as predicted, text differences were found such that as the variation in the type of information contained in the text increased, relationships with the structural variables increased.